Differential Depth Distribution of Microbial Function and Putative Symbionts Through Sediment-Hosted Aquifers in the Deep Terrestrial Subsurface
Total Page:16
File Type:pdf, Size:1020Kb
ARTICLES https://doi.org/10.1038/s41564-017-0098-y Differential depth distribution of microbial function and putative symbionts through sediment-hosted aquifers in the deep terrestrial subsurface Alexander J. Probst1,5,7, Bethany Ladd2,7, Jessica K. Jarett3, David E. Geller-McGrath1, Christian M. K. Sieber1,3, Joanne B. Emerson1,6, Karthik Anantharaman1, Brian C. Thomas1, Rex R. Malmstrom3, Michaela Stieglmeier4, Andreas Klingl4, Tanja Woyke 3, M. Cathryn Ryan 2* and Jillian F. Banfield 1* An enormous diversity of previously unknown bacteria and archaea has been discovered recently, yet their functional capacities and distributions in the terrestrial subsurface remain uncertain. Here, we continually sampled a CO2-driven geyser (Colorado Plateau, Utah, USA) over its 5-day eruption cycle to test the hypothesis that stratified, sandstone-hosted aquifers sampled over three phases of the eruption cycle have microbial communities that differ both in membership and function. Genome- resolved metagenomics, single-cell genomics and geochemical analyses confirmed this hypothesis and linked microorganisms to groundwater compositions from different depths. Autotrophic Candidatus “Altiarchaeum sp.” and phylogenetically deep- branching nanoarchaea dominate the deepest groundwater. A nanoarchaeon with limited metabolic capacity is inferred to be a potential symbiont of the Ca. “Altiarchaeum”. Candidate Phyla Radiation bacteria are also present in the deepest groundwater and they are relatively abundant in water from intermediate depths. During the recovery phase of the geyser, microaerophilic Fe- and S-oxidizers have high in situ genome replication rates. Autotrophic Sulfurimonas sustained by aerobic sulfide oxidation and with the capacity for N2 fixation dominate the shallow aquifer. Overall, 104 different phylum-level lineages are present in water from these subsurface environments, with uncultivated archaea and bacteria partitioned to the deeper subsurface. uch remains to be learned about how microbial communi overlooked due to sampling13 and primer bias13,20,21 and the spatial ties in the deep terrestrial subsurface vary with depth due variation in metabolic functions over depth transects including the Mto limited access to samples without contamination from deep subsurface (100 m below the ground) remains unexplored. drilling fluids or sampling equipment. Studies to date have anal Crystal Geyser is a coldwater, CO2driven geyser located geo ysed samples acquired by drilling1–3, from deep mines4,5, subsurface logically within the Paradox Basin, Colorado Plateau, Utah, USA22. research laboratories6,7 and sites of groundwater discharge8–11. These Originally an abandoned oil exploration well, the 800m deep verti investigations have shown that the terrestrial subsurface is popu cal borehole has served as a geyser conduit whose regular and sig lated by a vast array of previously undescribed archaea and bacteria. nificant flow rate (since 1936) provides uncontaminated access to At one site, an aquifer in Colorado (Rifle, USA), the diversity spans organisms present in underlying aquifers. Prior geological studies much of the tree of life12 and includes organisms of the Candidate have defined the region’s hydrostratigraphy, including the transmis Phyla Radiation (CPR)13, which may comprise more than 50% of sive Entrada, Navajo, Wingate and White Rim fractured sandstone all bacterial diversity14, and many other previously undescribed aquifers (listed in order of increasing depth), which are separated by bacterial lineages. Also present in the terrestrial subsurface are lowpermeability confining units23,24, through which limited verti previously unknown or little known archaea, including members cal connectivity for CO2, water and microbes is largely restricted of the DPANN (Diapherotrites, Parvarchaeota, Aenigmarchaeota, to faults and fractures25. A nearby research borehole provided fur Nanoarchaeota, Nanohaloarchaea)11,15, Altiarchaeum10, Lokiar ther geologic and aquifer geochemical information to 322 m below chaeota16 and Aigarchaeota17. ground surface26. Timeseries geochemical data collected over the A major question in subsurface microbiology relates to how ca. 5day eruption cycle suggest that Crystal Geyser is primarily organisms, and their capacities for carbon, nitrogen and sulfur sourced from the Navajo Sandstone, with increased contributions cycling, vary along depth transects through the terrestrial sub from the shallower Entrada Sandstone during major eruptions, and surface. Some evidence pointing to taxonomic variation between increased fraction of deeper water during minor eruptions26,27. 9 m and 52 m below the surface was obtained via a massive A survey of ribosomal proteins predicted from metagenome 16S ribosomal RNA gene survey at the Hanford Site18. Similar sequences from Crystal Geyser microbial communities revealed the variation and change of two functional genes were also detected for existence of a large phylogenetic diversity of previously unknown two shallow aquifers that were accessed via drilling in Germany19. bacteria and archaea in this system8, and a genomic resolution However, major groups of archaea and bacteria may have been study documented a high incidence of carbonfixation pathways9. 1Department of Earth and Planetary Science, University of California, Berkeley, CA, USA. 2Department of Geoscience, University of Calgary, Calgary, AB, Canada. 3Department of Energy Joint Genome Institute, Walnut Creek, CA, USA. 4Plant Development and Electron Microscopy, Department of Biology I, Biocenter LMU Munich, Planegg-Martinsried, Germany. Present address: 5Present address: Group for Aquatic Microbial Ecology, Biofilm Center, Department of Chemistry, University of Duisburg-Essen, Essen, Germany. 6Present address: Department of Plant Pathology, University of California, Davis, Davis, CA, USA. 7These authors contributed equally: Alexander J. Probst and Bethany Ladd. *e-mail: [email protected]; [email protected] 328 NatURE MICROBIOLOGY | VOL 3 | MARCH 2018 | 328–336 | www.nature.com/naturemicrobiology © 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. © 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved. NATURE MICROBIOLOGY ARTICLES A remaining question relates to the source regions and distribu level increased ~3.5 m over 33.5 h, potentially enabling microbes tions of these organisms. Here, we tracked the microbiology and the to thrive in microaerophilic boreholeaffected conditions. To sim associated geyser discharge geochemistry continuously through plify terminology, we henceforth refer to the source water composi out the full 5day geyser eruption cycle to test the hypothesis that tions as relatively ‘deep’ during the minor eruptions, ‘intermediate’ groundwater from stratified aquifers sampled at different stages (and boreholeaffected) during the recovery phase and ‘shallow’ of the cycle has microbial communities that differ in both mem during the major eruptions. Similar phase variations in the relative bership and function. Our analyses made use of a comprehensive depths of water composition were recently observed27. collection of more than 1,000 newly reconstructed genomes, both Analysis of microbial community composition in the 2015 bulk from metagenomes and single cells, as well as detailed physical and samples made use of a relatively comprehensive database of genomic chemical information that enabled linking of fluids to their ground information for the Crystal Geyser system (see Supplementary water source regions. Fig. 2 for sample processing and analysis overview). The genomic dataset included previously reported draft genomes from this system9 Results and new genomes reconstructed from sizefractionated samples Continuous in situ (downhole) monitoring of the geyser water obtained in April and October of 2014 (Supplementary Table 1). pressure throughout the field campaign defined the regular ~5day Samples included a post0.2 µ m fraction collected onto a 0.1µ m period of the eruption cycle (Supplementary Fig. 1). Sampling was filter to enrich for community members with ultrasmall cell sizes. conducted over a complete eruption cycle (24–29 May, 2015) dur Binning of assembled metagenomes from 27 different samples ing which microbial cells were continuously collected onto 0.1 µ m from five time points in 2014 used seven different algorithms filters. Time series of downhole temperature, electrical conductivity, (see Supplementary Methods) and resulted in 30,574 genomic total dissolved gas pressure and water samples (for major ion, trace bins (with multiple bins for the same genome generated by dif metal and dissolved gas analyses) were collected to associate specific ferent algorithms). Selection of the best quality bin generated for microorganisms with water from different geyser eruption intervals any organism in each sample resulted in 5,795 bins for bacteria and relative aquifer depths (Supplementary Fig. 1). and archaea, of which 2,216 were considered to be at least medium Time series of water pressure, electrical conductivity and tem quality (> 70% completeness based on singlecopy genes with less perature showed three Crystal Geyser eruption phases previously than three multiple singlecopy genes). After curation based on observed26,27: the recovery (relatively low water level, no eruptions, guaninecytosine content, coverage and taxonomy the database light CO2 bubbling), minor eruptions (short eruptions of